OpenAI is joining forces with South Korea’s Samsung and SK Hynix to secure advanced computer chips for its ambitious Stargate project.
The two Korean giants confirmed they will expand production to meet OpenAI’s rising demand for high-performance memory chips. These chips are critical for running OpenAI’s increasingly compute-hungry AI models at scale.
According to OpenAI, it could require as many as 900,000 semiconductor wafers every month from the companies. This massive order highlights the scale of its AI expansion.
According to the source[1], talks are also underway to build two AI data centers in South Korea. Samsung[2] is additionally studying floating data centers, which it believes could lower costs while also cutting carbon emissions.
The deals mark OpenAI’s latest push to secure the resources needed for growth. Alongside compute, the company is also focusing on power and cash to fuel its next phase.
Just last month, Nvidia revealed plans to invest up to $100 billion in OpenAI[3] as part of its expansion strategy. The partnership with Samsung and SK Hynix shows how OpenAI is strengthening its supply chain to keep pace with global AI demand.